Communications to the Editor
Vol. 9, No. 5, September-October 1976 fication I1 in the plane perpendicular to the fiber axis. Feature (2) suggests that in the b projection the disorder also has the same structure as modification 11, i.e., the @Of) behaves as if no disorder is present. On the other hand, poly(viny1idene fluoride) can take the planar zigzag conformation (modifications I and 111) in addition t o the (TGTE), conformation in modification I1 (T, trans; G and E, gauche and minus gauche). The following proposed structure is consistent with the above facts. A segment of a molecular chain in modification I1 changes to the planar zigzag conformation, for example, from . . . TGTGTGTC . . . t o . . . TGTGTTTTTGTG . . . . All the molecules in a crystallite take the same planar zigzag conformation a t the same level, forming a band (Figure 2). In the b projection, the planar zigzag part has the same structure as modification 11, though it deviates from modification I1 in the ac plane, and in the section perpendicular to the fiber axis it has the same periodicity as modification 11. This type of disorder is known as “kink bands” (Figure 3).3 The intensity distribution (called feature (3) above) was calculated on the basis of the theory for stacking fault^^,^ and found to agree with the observed intensity. This work will be published in detail elsewhere. The Bragg reflections from modification I1 containing the kink bands are not appreciably different from the usual ones. This suggests that the TGTG sequences are as long and regular as in the usual modification 11. On the other hand, the kink bands can be predicted to be short and disordered because there are some short intermolecular distances in the kink bands. Modification I1 transforms into the planar zigzag modifications by stretching (to modification I) and by heat treatment under high pressure (to modifications I and
planar zigzag ( T G T c)n
871
:I
Figure 3. Kink bands in modification I1 of poly(viny1idene fluoride) in the a projection.
It can be expected that the kink bands play an important role in these crystal transformation processes. References a n d Notes (1) Address correspondence to this author. ( 2 ) R. Hasegawa, Y. Takahashi, Y. Chatani, and H. Tadokoro, Polym. J . , 3,
600(1972). i31 For example. B. Wunderlich, “Macromolecular Physics”, Vol. 1, Academic
Press, New York, N.Y., 1973. Y.Komura, J . P h p SOC.J p n , 9, 169 (1954). (j)G . Allegra, Acta Crystallogr., 17,579 (1964). (6) R. Haspgawa, M. Kobayashi, and H. Tadokoro, Polym. J . , 3,591 (1972). ( 4 ) .J. Kakinoki and
Communications to the Editor Oligomerization Stereochemistry of Vinyl Monomers. 1. A Stereoselective Synthesis of meso-2,4-Di (2-pyridy1)pentane Studies of the oligomerization of vinyl derivatives, dienes, etc., are frequently of interest since they provide a simplified picture of the corresponding polymerization process. Thus, the stereochemistry of anionic oligomerization of dienes has been studied over the past several years.’ Though the anionic oligomerization of vinyl monomers has been e x p l ~ r e d ,the ~,~ stereochemistry of this process, to our knowledge, has not been reported. We now wish to report the addition of a-lithiated-2-ethylpyridine (1)to 2-vinylpyridine with sequential stereoselective methylation of the lithiated adduct (Scheme I). The reaction was carried out by a slow in vacuo distillation of 2-vinylpyridine onto a vigorously stirred T H F solution of the %ethylpyridyl Li salt that was in turn formed by metallation of 2ethylpyridine using n-BuLi. The methylation was similarly carried out by vacuum distillation of CH31onto the carbanion solution a t -78 “C. After workup of the reaction mixture, the product 3a was obtained by vacuum distillation, bp 106-108 “ C (0.15 mm). Product yield was about 60%. A small amount of polymer was also formed. The dimer stereochemistry was determined by 60 MHz NMR (meso-3a: 6(CH2) 1.77 and 2.25, a(CH) 2.80, b(CH3) 1.25 ppm; racemic-3a: 6(CH3) 1.19). Diastereomeric dimer mixtures were analyzed by determining
Scheme I
CHjCH,
1 -
CH3CH-Li+
I
‘1
+@
@ THF. - i 8 “C L@ 1
@‘@
THF. -i8 T
2
3 a = 2-pyridyl
b
=
4-pyridyl
the ratio of CH3 doublet absorptions. The product stereochemistry was found to be independent of the temperature of vinyl pyridine addition and time and temperature of storage of the dimer anion (2a). Hence, the stereoselection is most
872 Communications to the Editor
Macromolecules
Table I Methylation Stereochemistry of 1,3-Di(pyridyl)butane Anion as a Function of Temperature, Cation, Solvent, and Carbanion Structure Carbanion
Cation/solvent
Monomer
% meso of
T.O C
3a/3b
la la la la lb
Li/THF 2-VP -78 >99 Li/THF 2-VP 0 95 Li/THF, pyridinea 2-VP -78 a3 Rb/THF 2-VP -78 57 LiITHF 4-VP -78 -5oc a 50150 by volume. Carbanion salt was generated using amethylstyrene tetramer2-, Rbz+ salt. Measured by ratio of CH3 doublets; CH3 shifts of the diastereomers have not been identified. likely due to the methylation reaction. Equilibrium epimerization studies with potassium tert-butoxide in DMSO as catalyst showed that meso and racemic forms of 3 are of nearly equal stability, a result similar to that obtained by Flory and co-workers for the corresponding epimerization of 2,4-di~ h e n y l p e n t a n e The . ~ methylation stereochemistry thus appears to be kinetically controlled. A higher methylation temperature slightly decreases the stereoselectivity. The stereoselectivity of methylation is decreased when pyridine (50%by volume) was added after the formation of 2 (Table I). Also the use of Rb a-methylstyrene tetramer as a metallating agent significantly decreases the stereoselectivity of alkylation (Table I). These observations indicate that the high stereoselectivity is due to the strong coordinative ability of the Li ion. The above results are consistent with the presence of a strongly preferred conformation of the Li salt of 2a.5 If the Li ion is present above or below the planar carbanion,6 two diastereotopic ion pairs (4a and 4b) should exist, each of which
4a’
4b
II
It
M+
4a 4b
may occur in several conformations. The very high stereoselectivity of methylation can, in such a case, only be consistent with the predominant participation of either 4a or 4b in the reaction and an alkyl halide approach that is exclusively syn or anti with respect to metal cation. The pronounced effect of cation size and coordination on the stereochemistry of 3a strongly suggests Li ion coordination by the penultimate 2-pyridyl group. Several such conformations are possible for each ion pair diastereomer of which 4a‘ and 4b’ (corresponding to 4a and 4b, respectively) seem most reasonable. From inspection of models, 4a’ is expected to be more stable by several kilocalories on account of CH3-pyridine gauche and CH3-nitrogen lone pair interactions in 4b‘. Now,
the methylation reaction probably occurs cation side since the approach of the halide from the opposite side would result in the formation of energetically unfavorable product separated ion pairs. This cation side approach of halide also appears to be supported by recent work on stereoselective a l k y l a t i ~ n . ~ The predominance of the meso form in the methylation product of the Li salt of 2a is, therefore, consistent with the prevalence of 4a’. The above interpretation is supported by recent conductometric evidence8sgfor such 2-pyridyl Na+ ion coordination in species similar to 2a and by the fact that the methylation of the Li salt of 2b is not stereoselectivelo (Table I). Coordination of cation with the 4-pyridyl group of 2b seems impossible on the basis of inspection of models. The possible existence of two or more ion pair diastereomers is of interest in the stereochemistry of anionic polymerization processes.ll The role of ion pairing in the stereochemistry of anionic polymerization of vinyl monomers such as acrylates has been extensively studied but appears to be still incompletely understood. Most proposed descriptions of ion pair type processes focus on some type of coordination of cation with penultimate Lewis base substituent and incoming monomer.12 The above results tend to support such a model. The data also indicate that the anionic polymerization stereochemistry of 2-vinylpyridines and similar polar monomers may be affected by the relative proportion of the two “configurational” ion pair diastereomers of the carbanion site and by the direction of monomer approach with respect to cation. Oligomerization studies focusing on this aspect are presently underway. Acknowledgment. This work was supported by NSF through Grant DMR72-03130 References and Notes (1) See, for instance: F. Schue, Bull. SOC.Chim., 980 (1965);M. Morton and L. A. Falvo, Macromolecules, 6, 190 (1972);S. Bywater, D. J. Worsfold, and G. Hollingsworth, ibid., 5,389 (1972);A. Ulrich, Bull. SOC.Chim.,9, 2773 (1973). ( 2 ) C. L. Lee, J. Smid, and M. Szwarc, J . Phys. Chem., 66,904 (1962). (3) (a) L. Lochmann, M. Rodova, J. Petranek, and D. Lim, J . Polym. Sci., 12, 2295 (1974);(b) N. E. Sartoris and H. Pines, J. Org. Chem., 34,2119 (1969), and references therein. (4) A. D. Williams, J. I. Brauman, N. J. Nelson, and P. Flory, J . Am. Chem. Soc., 89,4807 (1967). (5) These 2- and 4-pyridyl substituted carbanion alkali salts have been shown to exist as contact ion pairs in T H F and dimethoxyethane: C. J. Chang, R. F. Kiesel, and T. E. Hogen-Esch, J. Am. Chem. SOC., 97, 2805 (1975). (6) CNDO calculations on picolyl carbanion alkali cation pairs carried out in our laboratory indicate that the cation is located above the carbanion plane as opposed to in the nodal plane. In agreement with these calculations, NMR studies on 2-ethylpyridyl carbanions indicate only a moderate charge delocalization into the pyridine ring (see ref 5 for references) so that a lithioenamine structure is not likely. Moreover, a nodal plane position of the Li ion cannot account for the observed stereoselectivity. (7) See, for instance: A. I. Meyers, G. Kraus, K. Kamata, and M. Ford, J. Am. Chem. Soc., 98,567 (1976);E. L. Eliel, A. Abatjoglou, and A. A. Hartmann, ibid., 94,4786 (1972); H. M. Walborsky and A. E. Young, ibid., 86,3288 (1964); T. Durst and M. Molin, Tetrahydron Lett., 1,63 (1975). (8) M. Fisher and M. Szwarc, Macromolecules, 3,23 (1970). (9) M. Tardi, D. Rouge, and P. Sigwalt, Eur. Polym. J. 3,85 (1967). (10) This recalls Natta’s observation that 2-vinylpyridines may be isotactically polymerized by organomagnesium compounds whereas 4-vinylpyridines cannot: G. Natta, G. Mazzanti, P. Longi, G. D. Asta, and F. Bernardini, J . Polym. Sci., 51,487 (1961). (11) B. D. Coleman and F. G. Fox, J . Am. Chem. SOC.,85,1241 (1963); B. D. Coleman and F. C. Fox, J . Chem. Phys., 38,1065 (1963). (12) See, for instance: D. J. Cram and K. R. Kopecky, J . Am. Chem. SOC.,81, 2748 (1959); W. E. Goode, F. H. Owens, and W. L. Meyers, J . Polym. Sci., 46,317 (1960);C. E. H. Bawn and A. Ledwith, Q. Reu., Chem. SOC.,16,361 (1962);J. Furakawa, Polymer, 3,87 (1962); W. Fowells, C. Schuerch, F. A. Bovey, and F. P. Hood, J . Am. Chem. Soc., 89,1396 (1967).
C. F. Tien and T. E. Hogen-Esch* D e p a r t m e n t of Chemistry, University of F l o r i d a , Gainesville, F l o r i d a 32611 Received A p r i l 5, 1976